Reduced plasma epidermal growth factor levels reflect poor glycaemic status in type 2 diabetes

Abstract

Background

Epidermal growth factor (EGF) plays a crucial role in cellular growth, differentiation, and pancreatic β-cell maintenance. Despite reports on EGF deficiency in diabetic animal models, its relevance in type 2 diabetes (T2D), particularly in relation to obesity, remains underexplored. The present study aimed to evaluate plasma EGF levels in individuals with and without T2D, assess its associations with glycaemic status and clinical parameters, and evaluate the influence of obesity on these relationships.

Methods

A total of 838 eligible participants were selected from the Kuwait Diabetes Epidemiology Program. Of those, 428 were included in a 1:1 case–control analysis (214 T2D and 214 non-diabetics). EGF was measured in plasma using ELISA. Associations between EGF with glycaemic and clinical variables were evaluated using Pearson’s correlation, multiple linear regression, and logistic regression analyses.

Results

Plasma EGF levels were significantly lower in individuals with T2D compared to non-diabetics (P < 0.001). Among non-diabetics, obese participants had a significantly lower EGF level than their non-obese counterparts (P = 0.03), while no such difference was observed in T2D. EGF negatively correlated with fasting blood glucose in both non-diabetics (P = 0.004) and T2D individuals (P < 0.001). In T2D, EGF negatively correlated with haemoglobin A1C (HbA1C) (P = 0.001), triglyceride (TG) (P = 0.021), and waist-to-hip ratio (WHR) (P = 0.014). Logistic regression confirmed that lower EGF levels were independently associated with T2D but not with general obesity (OR = 0.996, P = 0.001).

Conclusion

Reduced EGF levels are associated with poor glycaemic control in T2D. These findings highlight EGF’s potential as a biomarker for glycaemic dysregulation and support further investigation into its role in diabetes pathophysiology and complications.

Introduction

Diabetes refers to the hyperglycaemic state that occurs due to a historically classified group of metabolic illnesses, i.e. diabetes type 1 and 2 (1). Diabetes classification also includes hybrid forms of diabetes, other specific types (monogenic defects of β-cell function or insulin action, endocrine or exocrine diseases, drug or chemical induced, etc.), unclassified diabetes, and hyperglycaemia detected during pregnancy or gestational diabetes (2, 3). Aetiopathologically, diabetes develops due to disturbances in the secretion of insulin from pancreatic β-cells, the action of insulin, or both (3). In addition, defects in the metabolism of different food types (carbohydrates, fats, and proteins) can contribute to the development of diabetes (4).

Generally, diabetes causes nephropathic and neuropathic complications along with several characteristically occurring symptoms that usually always present in diabetic patients. These consist of polydipsia, polyuria, diplopia, weight loss, and susceptibility to infections (4). Type 2 diabetes (T2D) symptoms, however, are usually mild and can be hard to detect/diagnose, caused by its prolonged progression; compared to other types of diabetes, its complications usually present at diagnosis, adding more complexity to its management and treatment (5).

Essentially, T2D is a metabolic disease that involves complex changes in the endocrine milieu, resulting primarily from disruptions in insulin and glucose homoeostasis (6). Its global burden is rising rapidly, with projections estimating over 600 million cases worldwide by 2040 (7). In Kuwait, the prevalence of T2D among adults is alarmingly high, exceeding 19% (8). Despite advancements in therapeutic options, glycaemic control remains suboptimal for many patients, highlighting the need for novel biomarkers that can better reflect disease status and progression.

T2D develops due to insulin resistance and/or insulin insufficiency, which is largely due to the reduction in the number of pancreatic β-cells or their increased apoptosis (9). Insulin insufficiency leads to persistent hyperglycaemia, which is a major cause of diabetic complications (10). Inadequate glycaemic control directly contributes to the occurrence and progression of diabetic-related microvascular and macrovascular complications (11).

Recently, there has been a growing interest in understanding the role of growth factors on the prevention and treatment of diabetic complications. Growth factors refer to endogenously produced polypeptides that regulate various cellular functions, including cell growth, proliferation, differentiation, and tissue repair (12). Several studies have shown that growth factor signalling is affected in both obesity and T2D (13, 14). Excess growth factor production occurs in tissues where fibrosis is predominant, as observed in diabetic kidney disease, while there is a reduction in growth factor production in tissues where inflammation is predominant, as in diabetic foot ulcers.

Epidermal growth factor (EGF) is a low-molecular-weight polypeptide originally identified for its potent mitogenic effects on epithelial tissues. EGF plays a pivotal role in promoting cellular proliferation, survival, migration, and differentiation across multiple organ systems (15, 16). Within the pancreas, EGF supports β-cell proliferation, differentiation, and regeneration, processes that are critical for maintaining insulin secretion and glucose homoeostasis (17, 18). Experimental models have shown that exogenous EGF administration can stimulate β-cell mass expansion and improve glycaemic control, underscoring its potential therapeutic relevance in diabetes (19, 20).

Alterations in EGF expression and signalling have been implicated in various diabetic complications. In animal models of diabetes, reduced EGF levels have been linked to impaired pancreatic regeneration, wound healing deficits, and nephropathy (13, 21, 22). Moreover, EGF has been shown to modulate lipid metabolism by regulating apolipoprotein secretion (23, 24) and influencing adipocyte function, suggesting that it could be involved in broader metabolic processes beyond glycaemic regulation. Interestingly, EGF concentrations decline with age and vary by sex (25, 26), factors that could confound their associations with disease states and must be considered in clinical studies.

Despite the biological plausibility, few studies have systematically examined circulating plasma EGF levels in human T2D populations, particularly in the context of obesity. Obesity, especially visceral fat accumulation, is known to exacerbate insulin resistance and systemic inflammation, potentially impacting growth factor signalling (27). Understanding whether EGF levels are altered in T2D and how they relate to clinical and biochemical markers may reveal new insights into metabolic dysregulation. The present study aimed to assess plasma EGF concentrations in individuals with and without T2D in a Kuwaiti cohort. We further explored the relationship between EGF and key clinical parameters, while considering the influence of obesity.

Materials and methods

Study design

Initially, a total of 990 participants were recruited from October 2021 to February 2022 as part of the Kuwait Diabetes Epidemiology Program (KDEP), a population-based initiative designed to investigate metabolic and cardiovascular risk factors among individuals across the glycaemic spectrum. Inclusion criteria were participants aged 18 years or older. The exclusion criteria were as follows: pregnancy and a history of chronic illness or malignancy and morbid obesity body mass index (BMI ≥ 40 kg/m²) to reduce heterogeneity and potential confounding, given the profound metabolic and inflammatory alterations and comorbidities associated with morbid obesity. In addition, the lower BMI inclusion threshold was set at 19.5 kg/m² to ensure exclusion of underweight individuals (BMI < 18.5 kg/m²), in accordance with standard BMI classifications (28). The total number of eligible participants post-exclusion criteria was 838 (Fig. 1). The study cohort included 469 non-diabetic controls and 369 T2D individuals. All volunteers were subjected to pre-screening to determine their eligibility to participate in the study. T2D was defined by fasting blood glucose (FBG) ≥ 7 mmol/L under treatment, or self-reports of previously diagnosed T2D. The diagnosis of diabetic participants was based on American Diabetes Association criteria (29).

Figure 1.

Flowchart representing details of the participant recruitment process for the study. T2D: type 2 diabetes.

Based on previous findings showing that EGF levels vary with age and gender (25, 26), case–control matching was implemented to facilitate unbiased comparisons of EGF levels between study groups. Participants with T2D were individually matched 1:1 with normoglycemic controls based on gender and age (±5 years) using a case–control design. Eventually, 214 non-diabetic controls and 214 T2D were included in the final analysis. Participants were stratified according to BMI and were classified as either non-obese (19.5 ≤ BMI < 30 kg/m²) or obese (30 ≤ BMI < 40 kg/m²). Based on this stratification, the non-diabetic group included 161 non-obese and 53 obese participants and the T2D group included 131 non-obese and 83 obese participants.

This study was approved by the Ethical Committee at the Dasman Diabetes Institute (DDI) (Project No.: RA-HM-19-030) and was performed in accordance with the principles of the Declaration of Helsinki, as revised in 2008. Written informed consent was obtained from all study participants.

Anthropometric and biochemical measurements

Fasting blood samples were collected into vacutainer-EDTA tubes and centrifuged at 400 g for 10 min to separate the plasma, which was aliquoted and stored at −80°C until assays. FBG, triglyceride (TG), total cholesterol (TC), low-density lipoprotein (LDL), and high-density lipoprotein (HDL) were measured with a Siemens Dimension RXL chemistry analyser (Diamond Diagnostics, USA). Glycated haemoglobin (HbA1c) levels were measured using the Variant device (Bio-Rad Laboratories, USA). The insulin levels were quantified using the Access Insulin Assay (Beckman Coulter, USA).

Measurement of EGF plasma levels using enzyme-linked immunosorbent assay (ELISA)

EGF plasma levels were determined using the DuoSet ELISA kit (DY236, R&D Systems, UK). Frozen plasma samples were thawed on ice and centrifuged at 10,000 g for 5 min at 4°C to remove any debris or platelets. ELISA was performed according to the manufacturer’s instructions. Samples were analysed randomly to prevent bias. The intra-assay coefficient of variation was 4.45% at 14.5 pg/mL (n = 20) and 3.23% at 80 pg/mL (n = 20), and the inter-assay coefficient of variation was 6.1% at 15 pg/mL (n = 20) and 5.6% at 78 pg/mL (n = 20). The detection limit was 3.9 pg/mL, and the range of linearity was set at 3.9–500 pg/mL.

Statistical analysis

The statistical analysis was performed using the GraphPad Prism software (USA) and SPSS software (IBM SPSS, USA). Given that all statistical tests used in the present study require data normality, continuous variables were normalized using the inverse distribution function in SPSS. Student’s t-test was used to determine significance between two groups, while one-way ANOVA was used for comparison between three groups. Multiple linear regression analysis models (crude or adjusted for age, gender, and BMI) were implemented to identify parameters independently associated with EGF. The association of EGF with T2D and obesity was analysed using multivariate logistic regression models with crude or adjusted odds ratio. Pearson’s partial correlation adjusted for age and gender was used to evaluate associations between plasma EGF levels and study variables. Data are presented as mean ± standard deviation (SD). All P-values reported were two-tailed, and P-value< 0.05 was considered significant.

Results

Characteristics of the study population

Table 1 presents the demographic and clinical characteristics of the full study cohort, comprising 469 non-diabetic and 369 T2D participants. The T2D group had a significantly higher proportion of males (P < 0.001) and was significantly older than the non-diabetic group (P < 0.001). BMI and waist–hip ratio (WHR) were also significantly elevated in the T2D group (both P < 0.001). As expected, T2D individuals had significantly higher levels of FBG, HbA1c, and insulin (all P < 0.001). TG levels were significantly elevated (P < 0.001), while HDL cholesterol was lower in the T2D group (P < 0.001). There were no significant differences in TC (P = 0.932) or LDL cholesterol (P = 0.609) between groups. CRP was higher in the T2D group (P = 0.025).

Table 1.

Demographics and characteristics of the full study cohort.

Phenotype	Non-diabetics	T2D	P-value
n	469	369
Gender (M/F)	221/248	292/77	<0.001 *
Age (years)	38.38 ± 8.88	50.35 ± 10.43	<0.001†
BMI (Kg/m2)	27.92 ± 6.04	30.67 ± 5.88	<0.001†
Waist–hip ratio	0.89 ± 0.09	0.94 ± 0.06	<0.001†
FBG (mmol/L)	4.88 ± 2.10	7.74 ± 2.41	<0.001†
HbA1C (%DCCT)	5.22 ± 1.24	7.07 ± 1.70	<0.001†
Insulin (mU/L)	8.0 ± 7.43	11.61 ± 8.40	<0.001†
TC (mmol/L)	5.12 ± 1.00	5.11 ± 1.06	0.932†
TG (mmol/L)	1.31 ± 1.16	1.91 ± 1.11	<0.001†
HDL (mmol/L)	1.24 ± 0.39	1.07 ± 0.39	<0.001†
LDL (mmol/L)	3.29 ± 0.89	3.25 ± 0.95	0.609†
CRP (mg/dL)	3.69 ± 3.79	4.30 ± 3.40	0.025†

Bold indicates statistical significance (P < 0.05). BMI, body mass index; CRP, c-reactive protein; FBG, fasting blood glucose; HbA1C, glycated haemoglobin; HDL, high-density lipoprotein; LDL, low-density lipoprotein; T2D, type 2 diabetes; TC, total cholesterol; and TG, triglycerides.

^*Chi-squared test.

^† Student’s t-test.

The case–control matched cohort included 214 T2D individuals and 214 non-diabetic controls (Table 2). There were no significant differences in age or gender distribution between groups. Compared to the non-diabetic group, participants with T2D had significantly higher BMI, WHR, and CRP levels (P < 0.001, P = 0.002, and P = 0.038, respectively). In terms of glycaemic status, T2D individuals had markedly elevated FBG, HbA1c, and insulin levels relative to controls (all P < 0.001). Lipid profiles showed significant differences in TG and HDL levels between groups, with higher TG and lower HDL in the T2D group (both P < 0.001), while TC and LDL did not differ significantly.

Table 2.

Demographics and characteristics of the case–control matched study cohort.

Phenotype	Non-diabetics	T2D	P-value
n	214	214
Gender (M/F)	153/61	153/61	1.000*
Age (years)	44.10 ± 8.51	45.46 ± 9.18	0.113†
BMI (Kg/m2)	27.64 ± 4.68	29.47 ± 4.91	<0.001†
Waist–hip ratio	0.91 ± 0.06	0.93 ± 0.07	0.002†
FBG (mmol/L)	4.78 ± 2.03	7.64 ± 2.60	<0.001†
HbA1C (%DCCT)	5.26 ± 1.08	6.83 ± 1.86	<0.001†
Insulin (mU/L)	8.29 ± 7.57	11.18 ± 8.37	<0.001†
TC (mmol/L)	5.24 ± 0.98	5.22 ± 0.99	0.862†
TG (mmol/L)	1.46 ± 1.03	1.95 ± 1.12	<0.001†
HDL (mmol/L)	1.22 ± 0.36	1.07 ± 0.39	<0.001†
LDL (mmol/L)	3.41 ± 0.87	3.33 ± 0.91	0.358†
CRP (mg/dL)	3.58 ± 3.66	4.36 ± 4.03	0.038†

Bold indicates statistical significance (P < 0.05). BMI, body mass index; CRP, c-reactive protein; FBG, fasting blood glucose; HbA1C, glycated haemoglobin; HDL, high-density lipoprotein; LDL, low-density lipoprotein; T2D, type 2 diabetes; TC, total cholesterol; and TG, triglycerides.

^*Chi-squared test.

^† Student’s t-test.

Table 3 presents the demographic and clinical characteristics of non-diabetic and T2D participants stratified by obesity status. In the non-diabetic group, gender distribution did not differ significantly between obese and non-obese participants (P = 0.507), and age was also comparable between groups. Obese participants exhibited a significantly higher BMI (P < 0.001) and WHR (P = 0.021). Insulin levels were significantly higher in obese individuals (P < 0.001), and TC was modestly but significantly elevated (P = 0.033). No significant differences were observed for TG, HDL, LDL, FBG, or HbA1c, while CRP showed a borderline, non-significant trend towards higher levels in obese participants (P < 0.10). In contrast, within the T2D group, a significant difference in gender distribution was observed between obese and non-obese participants (P < 0.001). Obese individuals also had a significantly higher BMI (P < 0.001). No significant differences were detected for age, WHR, FBG, HbA1c, or insulin levels. However, obese participants exhibited significantly higher TG (P = 0.001), TC (P < 0.001), and LDL levels (P = 0.003), along with elevated CRP levels (P = 0.007).

Table 3.

Demographics and characteristics of the case–control matched study cohort stratified by BMI.

Population	Non-diabetics			T2D
Phenotype	Non-obese	Obese	P-value	Non-obese	Obese	P-value
n	161	53		131	83
Gender (M/F)	117/44	36/17	0.507	107/24	46/37	<0.001*
Age (years)	43.30 ± 9.14	44.41 ± 10.38	0.457†	45.80 ± 10.4	47.41 ± 9.66	0.259†
BMI (Kg/m2)	25.61 ± 3.29	33.77 ± 2.37	<0.001†	26.44 ± 3.44	34.26 ± 2.46	<0.001†
Waist–hip ratio	0.91 ± 0.06	0.93 ± 0.08	0.043†	0.93 ± 0.06	0.93 ± 0.07	0.867†
FBG (mmol/L)	4.7 ± 2.05	5.04 ± 1.20	0.286†	7.51 ± 2.66	7.84 ± 2.50	0.372†
HbA1C (%DCCT)	5.24 ± 1.08	5.32 ± 1.09	0.642†	6.96 ± 1.64	6.61 ± 2.15	0.176†
Insulin (mU/L)	6.20 ± 7.17	12.33 ± 7.38	<0.001†	10.36 ± 8.38	12.47 ± 8.24	0.073†
TC (mmol/L)	5.18 ± 0.10	5.42 ± 0.94	0.111†	5.02 ± 0.99	5.54 ± 0.91	<0.001†
TG (mmol/L)	1.43 ± 1.07	1.54 ± 1.13	0.521†	1.75 ± 1.08	2.26 ± 1.12	0.001†
HDL (mmol/L)	1.22 ± 0.36	1.23 ± 0.34	0.856†	1.09 ± 0.37	1.05 ± 0.42	0.427†
LDL (mmol/L)	3.35 ± 0.89	3.58 ± 0.80	0.105†	3.18 ± 0.91	3.56 ± 0.85	0.003†
CRP (mg/dL)	3.34 ± 3.70	4.3 ± 3.47	<0.100†	3.77 ± 4.02	5.27 ± 3.89	0.007†

BMI, body mass index; CRP, c-reactive protein; eGFR, estimated glomerular filtration rate; FBG, fasting blood glucose; HbA1C, glycated haemoglobin; HDL, high-density lipoprotein; LDL, low-density lipoprotein; T2D, type 2 diabetics; TC, total cholesterol; and TG, triglycerides.

^*Chi-squared test.

^† Student’s t-test.

Determinants of EGF predictors in the entire study population

To examine the association between EGF levels, T2D status, obesity, age, and gender, we performed multiple linear regression models using plasma EGF as the dependent variable (Table 4). Model 1 included only T2D status as a predictor (unadjusted analysis). Model 2 was adjusted for age and gender, while model 3 was adjusted for age and gender and BMI. The inclusion of BMI allowed the assessment of its independent contribution and its effect on the T2D–EGF relationship. Model 2 was statistically significant (F_3,831 = 14.316, P < 0.001, r² = 4.6%), where T2D status was a significant negative predictor of EGF (β = −0.119, P = 0.002), indicating that T2D individuals had lower EGF levels. As expected, age was also negatively associated with EGF (β = −0.123, P < 0.001), while gender was not a significant predictor (β = 0.050, P = 0.160). In Model 3, the addition of BMI increased the adjusted r² from 4.6 to 5.8% (F_4,829 = 12.655, P < 0.001, r² = 5.8%). The negative association between T2D status and EGF remained statistically significant but slightly attenuated (β = −0.103, P = 0.008). BMI emerged as an independent negative predictor of EGF (β = −0.099, P = 0.005), highlighting its possible contribution to EGF variation.

Table 4.

Multiple regression analysis to identify parameters associated with EGF.

Model	Model*		Model†		Model‡
EGF	β	P-value	β	P-value	β	P-value
T2D	−0.185	<0.001	−0.119	0.002	−0.103	0.008
Age	–	–	−0.123	<0.001	−0.101	0.007
Gender	–	–	0.050	0.160	0.054	0.129
BMI	–	–	–	–	−0.099	0.005

^*Model: crude.

^† Model: included age and gender.

^‡ Model: adjusted for age, gender, and BMI.

Bold indicates statistical significance (P < 0.05). BMI, body mass index; EGF, epidermal growth factor; and T2D, type 2 diabetes.

To further confirm these findings, plasma EGF levels were compared in the entire study cohort stratified into three age groups (<35, 35–50, and >50 years) based on the age distribution of the population (mean). A significant decline in plasma EGF levels with increasing age (age < 35 = 97.44 ± 86.39 pg/mL ± SD, n = 204; 35–50 = 78.53 ± 70.23 pg/mL ± SD, n = 414; and age > 50 = 64.12 ± 56.05 pg/mL ± SD, n = 220) (P < 0.001) (Fig. 2A). Furthermore, females (84.04 ± 68.18 pg/mL ± SD, n = 325) (P < 0.01) had significantly higher EGF levels than males (72.59 ± 64.30 pg/mL ± SD, n = 513) (P < 0.01) (Fig. 2B). These findings align with previous research, indicating that EGF concentrations decrease with age and may differ between sexes (20, 21). Therefore, age and sex were identified as important variables that could bias comparisons across groups. The decision to match on these factors ensured that any observed differences in EGF levels between clinical groups were not attributable to imbalances in age or gender distribution. Although BMI emerged as a significant negative predictor of EGF in the regression analysis (Table 4), natural variation in BMI was reserved to allow for secondary analyses involving obesity status.

Figure 2.

Circulating plasma levels of EGF in the study population categorized by age and gender. (A) The population was stratified based on age (<35 (n = 204), 35–50 (n = 414), and >50 years (n = 220)). (B) The population was stratified based on gender (male (n = 513) and female (n = 325)). The circles represent individual values, the boxes represent means, and the deviations represent standard deviations. *P < 0.05, ***P < 0.001.

Correlation analysis between EGF and glycaemic and other clinical variables

Pearson’s correlation coefficient analysis, adjusted for age and gender in non-diabetics and T2D individuals, was performed to assess the relationship between circulating EGF levels and glycaemic status and other clinical parameters (Table 5). In the non-diabetic group, EGF was negatively correlated with FBG (r = −0.195, P = 0.004) and positively correlated with CRP (r = 0.151, P = 0.029), although no other significant associations were observed. In contrast, among individuals with T2D, EGF showed a stronger negative correlation with glycaemic and other study variables. This included FBG (r = −0.335, P < 0.001), HbA1c (r = −0.220, P = 0.001), TG (r = −0.159, P = 0.021), and WHR (r = −0.169, P = 0.014) (Table 5). No significant correlations were observed between EGF and insulin, HDL, or LDL in either group. Collectively, these results indicate that in individuals with T2D, lower EGF levels are associated with poor glycaemic control, hypertriglyceridemia, and central adiposity.

Table 5.

Correlation analysis between circulating EGF levels and physical, clinical, and biochemical parameters in non-diabetic and T2D participants, adjusted for age and gender.

Phenotype	EGF (pg/mL)
	Non-diabetics		T2D
	r	P-value	r	P-value
BMI (Kg/m2)	−0.087	0.205	−0.123	0.074
Waist–hip ratio	0.094	0.172	−0.169	0.014
FBG (mmol/L)	−0.195	0.004	−0.335	<0.001
HbA1C (%DCCT)	−0.033	0.629	−0.220	0.001
Insulin (mU/L)	0.033	0.630	−0.039	0.575
TC (mmol/L)	−0.072	0.298	−0.032	0.647
TG (mmol/L)	0.077	0.263	−0.159	0.021
HDL (mmol/L)	0.013	0.856	0.085	0.220
LDL (mmol/L)	0.039	0.571	−0.017	0.809
CRP (mg/dL)	0.167	0.016	−0.088	0.205

Bold indicates statistical significance (P < 0.05). BMI, body mass index; CRP, c-reactive protein; EGF, epidermal growth factor; FBG, fasting blood glucose; HbA1C, glycated haemoglobin; HDL, high-density lipoprotein; LDL, low-density lipoprotein; T2D, type 2 diabetes; TC, total cholesterol; and TG, triglycerides.

Association between EGF levels with type 2 diabetes and obesity

In this case–control cohort, the plasma levels of EGF were significantly lower in individuals with T2D, as compared to non-diabetics (P < 0.001) (Fig. 3A). After further classifying the population based on BMI, obese non-diabetics exhibited significantly lower EGF levels compared to their non-obese counterparts (P = 0.03) (Fig. 3B). On the other hand, EGF levels did not significantly differ between non-obese and obese individuals in the type 2 diabetic group (P = 0.949) (Fig. 3C).

Figure 3.

Circulating levels of EGF in the case–control matched cohort. (A) Plasma levels of EGF in T2D participants (n = 214) relative to non-diabetics (n = 214). (B) Plasma levels of EGF in obese non-diabetics (n = 161) relative to non-obese non-diabetics (n = 53). (C) Plasma levels of EGF in obese T2D individuals (n = 131) relative to non-obese T2D participants (n = 83). The circles represent individual values, the boxes represent means, and the deviations represent standard deviations. *P < 0.05, ***P < 0.001.

To determine whether the diabetes-associated reduction in EGF was modified by obesity, we assessed changes in EGF levels between T2D individuals and non-diabetics within each BMI subgroup. As shown in Fig. 4A, among non-obese individuals, those with T2D had significantly lower plasma EGF levels compared to non-diabetic participants (P < 0.001). In contrast, no significant difference in EGF levels was observed between T2D and non-diabetic individuals within the obese subgroup (Fig. 4B). These results suggest that the diabetes-associated reduction in EGF is more pronounced in non-obese individuals, whereas in the context of obesity, this difference may be attenuated or masked. To further investigate whether plasma EGF levels were predictive of T2D or obesity, multivariate logistic regression analyses were performed (Table 6). Across all models, lower EGF levels were significantly associated with increased odds of having T2D. In the unadjusted model (model 1), EGF was significantly associated with T2D (OR = 0.995, 95% CI: 0.993–0.998, P < 0.001), and this association remained robust after adjustment for age and gender (model 2: OR = 0.996, 95% CI: 0.993–0.999, P < 0.001) and in the fully adjusted model including BMI (model 3: OR = 0.996, 95% CI: 0.993–0.999, P = 0.001).

Figure 4.

Impact of obesity on circulating EGF levels in the case–control matched cohort. (A) Plasma levels of EGF in non-obese T2D participants (n = 131) relative to non-obese non-diabetics (n = 161). (B) Plasma levels of EGF in obese T2D participants (n = 83) relative to obese non-diabetics (n = 53). The circles represent individual values, the boxes represent means, and the deviations represent standard deviations. ***P < 0.001.

Table 6.

Multivariate logistic regression analysis predicting the likelihood of T2D and obesity based on EGF plasma levels.

	T2D			Obesity
	OR	95% CI	P-value	OR	95% CI	P-value
Model*	0.995	0.993–0.998	<0.001	0.997	0.995–1.000	0.058
Model†	0.996	0.993–0.999	<0.001	0.997	0.995–1.000	0.065
Model‡	0.996	0.993–0.999	0.001	–	–	–

^*Model: crude.

^† Model: included age and gender.

^‡ Model: adjusted for age, gender and BMI.

Bold indicates statistical significance (P < 0.05). OR, odds ratio; CI, confidence interval; and T2D, type 2 diabetes.

In contrast, EGF was not significantly associated with obesity in any of the models.

Predictive performance of EGF for T2D

To further evaluate the predictive performance of circulating EGF for T2D status, ROC curve analysis was performed in the age- and sex-matched cohort. Circulating EGF demonstrated a modest but statistically significant discriminative ability (AUC = 0.595, 95% CI: 0.542–0.649, P < 0.001). The optimal cut-off value identified using the Youden index was 42.65 pg/mL, corresponding to a sensitivity of 45% and a specificity of 72% (Fig. 5).

Figure 5.

ROC curve analysis performed to identify the cut-off value of EGF as a biomarker for T2D. AUC for EGF (0.595 (0.542–0.649), P < 0.001).

Discussion

In our study, we demonstrated the potential role of EGF as a viable indicator of glycaemic status in T2D. We demonstrated that age, gender, and T2D status were associated with circulating plasma EGF concentrations, and EGF levels were significantly reduced in T2D individuals. These findings were consistent with previous reports in animal models (17, 21, 22) and human studies exhibiting reduced EGF mRNA expression in blood samples from individuals with T2D (30).

In pancreatic tissue, EGF contributes to β-cell proliferation, maintenance, and regeneration (15, 31); therefore, deficiency of EGF in T2D may reflect reduced β-cell mass or impaired insulin production, both hallmarks of the disease (9). Subsequently, we observed a negative correlation between EGF and glycaemic markers (FBG and HbA1c) in T2D individuals, suggesting a potential regulatory role for EGF in glucose metabolism. This aligns with in vitro and in vivo studies showing that EGF supplementation enhances insulin production and corrects hyperglycaemia (19, 20). Furthermore, a recent study reported a negative correlation between serum EGF and both FBG and HbA1c in T2D individuals with controlled and uncontrolled diabetes (32). Interestingly, this study reported a negative correlation between serum EGF and homoeostatic model assessment for insulin resistance (HOMA-IR). Interpretation of circulating EGF levels in T2D should also consider the potential influence of diabetes-related therapies. Glucose-lowering agents, such as GLP-1 receptor agonists, SGLT2 inhibitors, and insulin, and lipid-lowering therapies, including statins, are known to modulate glycaemic control, inflammatory status, and metabolic pathways that may intersect with EGF signalling. In addition, differences in measurement methods and the body’s own compensatory mechanisms adjusting other hormonal levels could mask potential correlations.

Our study also found a negative correlation between EGF and TG in individuals with T2D. These observations suggest a broader role for EGF in lipid metabolism. The association of EGF with lipid metabolism is thought to be via EGF’s regulation of production of apolipoproteins A and B in human fetal and colon cells (23, 33). In Caco-2 cells, EGF inhibited the secretion of apolipoprotein B, which is a constituent of very-low-density lipoprotein (VLDL), the major carrier of TGs in the blood (24). The inhibitory effect of EGF on apolipoprotein B secretion has been hypothesized to contribute towards wound healing, perhaps by diverting metabolic processes away from TG secretion. Therefore, our results suggest a protective role for EGF against dyslipidaemia.

Interestingly, the diabetes-associated reduction in EGF was evident in non-obese T2D individuals but not in their obese counterparts. This suggests that obesity may attenuate or mask the effect of diabetes on EGF levels. One possible explanation is that obesity, independent of T2D, contributes to a pro-inflammatory and metabolically stressed state that already suppresses EGF production. As a result, the additional impact of diabetes on EGF may become less detectable in the presence of excess adiposity. These findings highlight the complex interplay between fat distribution, glycaemic status, and growth factor regulation and underscore the importance of stratifying analyses by obesity when evaluating biomarker behaviour in metabolic disease. However, we observed a significant negative correlation between EGF and WHR in T2D individuals, suggesting that central adiposity, rather than overall body size, may be more closely linked to reduced EGF levels in T2D. Furthermore, obese non-diabetic individuals exhibited lower EGF levels compared to their non-obese counterparts. This could be explained by the higher WHR in the non-diabetic obese subgroup. This is interesting as WHR is a known indicator of visceral fat accumulation, which is metabolically active and associated with chronic low-grade inflammation (27). In clinical populations and large-scale epidemiological studies, WHR has been positively associated with elevated levels of high-sensitivity CRP (hs-CRP), independent of BMI (34, 35). Although no correlation was found between EGF and CRP in T2D, possibly due to the use of a standard CRP assay rather than a high-sensitivity method, the observed association between EGF and WHR warrants further investigation.

Importantly, our study demonstrated that circulating EGF levels, independent of age and sex, were associated with an increased T2D risk but not obesity in logistic regression analysis, highlighting a robust relationship between EGF and glycaemic status. Consistent with this association, ROC analysis indicated a modest discriminative performance. The sensitivity and specificity values suggest that EGF alone is unlikely to function as a stand-alone diagnostic marker but may contribute as part of a multi-marker panel reflecting glycaemic dysregulation.

Importantly, although reduced circulating EGF levels were associated with T2D in the present study, the therapeutic modulation of EGF signalling remains complex and tissue dependent. In the kidney, excessive activation of the EGF receptor (EGFR) has been linked to renal hypertrophy, fibrosis, and progression of diabetic kidney disease, suggesting that sustained or unregulated EGF signalling may be detrimental in this context. Experimental evidence indicates that EGFR inhibition can attenuate renal injury and metabolic dysfunction in diabetes, highlighting the dual and context-specific roles of EGF signalling in metabolic disease (36). Therefore, while our findings support a relationship between reduced circulating EGF levels and glycaemic dysregulation, any therapeutic strategies targeting EGF would require careful consideration of tissue specificity, signalling balance, and disease stage rather than systemic EGF supplementation.

Overall, reduced EGF levels may reflect impaired glycaemic regulation in T2D, and further longitudinal and mechanistic studies are warranted to determine whether modulation of EGF signalling could contribute to improved glycaemic control and reduced risk of diabetic complications.

Limitations

Despite the well-characterized participants and case–control matching, several limitations should be acknowledged. First, although many of the observed correlations between EGF and clinical parameters reached statistical significance, caution should be exercised when interpreting the clinical relevance of individual correlation coefficients. Second, the unequal distribution between BMI subgroups, particularly the smaller number of obese participants, may have limited the power to detect meaningful differences or interactions with EGF. Third, future studies with larger and more gender-balanced cohorts are needed to confirm and expand on these observations. Finally, as medication data were not uniformly available, adjustment for these factors was not possible. Commonly prescribed diabetes-related therapies, including glucose-lowering and lipid-modifying agents, may influence metabolic and inflammatory pathways that intersect with EGF signalling. Therefore, residual confounding by treatment exposure cannot be excluded and should be considered when interpreting the observed associations.

Conclusion

This study provides evidence that circulating plasma EGF levels are significantly reduced in T2D individuals. Through a case–control design that matched for age and gender, we demonstrated that lower EGF levels are independently associated with poorer glycaemic control. The lack of association between EGF and BMI, coupled with a significant negative correlation between EGF and WHR in individuals with T2D, suggests that central adiposity, rather than overall obesity, may be more strongly associated with reduced EGF levels. Overall, our findings support the utility of plasma EGF as a potential biomarker of glycaemic dysregulation and systemic inflammation in T2D. Future longitudinal and mechanistic studies are warranted to explore whether EGF supplementation or modulation can mitigate metabolic deterioration and reduce the risk of diabetic complications.

Declaration of interest

The authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of the work reported.

Funding

This work did not receive any specific grant from any funding agency in the public, commercial or not-for-profit sector.

Author contribution statement

Y Alshammari performed the literature search, designed the study, performed ELISA and statistical analysis, interpreted the results, and wrote the original draft. R Alshammari performed statistical analysis. R Nizam was involved in sample processing and storage and data interpretation. MB Assas, M Assas, and H Ali interpreted the data and critically revised the manuscript. L Koti wrote, reviewed, and edited the manuscript. F Al-Mulla collected clinical data, performed validation, supervised the study, and critically revised the manuscript. All authors read and approved the final manuscript.

Data availability

The data generated in the present study may be requested from the corresponding author.

Consent for publication

This study was approved by the Ethical Committee at the Dasman Diabetes Institute (DDI) (Project No.: RA-HM-19-030) and was performed in accordance with the principles of the Declaration of Helsinki, as revised in 2008. Written informed consent was obtained from all study participants.